You are on page 1of 19

Corruption in Japan An Economists Perspective Werner Pascha*

1. Introduction Since the years of the bubble economy and the ensuing economic and political turmoil after the bubble burst, political corruption in Japan has become a major concern for the general public, for the political and business elite and for scholars interested in Japan. Most of the analyses undertaken are written from a political (science) point of view, and they quite naturally usually focus on the developments in Japan (Hirose 1993, Mitchell 1996, Blechinger 1998).

In this paper I take a different approach. As an economist, I try to deal with the subject area employing methods and explanatory ideas taken from economics, without being a specialist on corruption myself. Moreover, I try to adopt some comparative perspective in order to see to what extent Japan is typical or atypical (special or unique). Such a comparative perspective is found in many economic studies on corruption, because a lot of the literature is interested in how corruption influences growth and development.

Although the contribution which this paper can make may be rather limited, I hope that to some extent it is programmatic for what Japanese Studies should strive for in the future. One aspect is to shed light on specific issues from various disciplinary points of view. This can help

This paper is based on my remarks as a commentator on the panel Shadow Politics Political Corruption in Japan at the annual meeting of the Association for Asian Studies in Boston, March 14, 1999, organised and chaired by Verena Blechinger, German Institute for Japanese Studies, Tokyo. Although the papers presented during the session were drafts and, therefore, cannot be quoted here, it is obvious that I very much profited from them. In particular I would like to acknowledge: Verena Blechinger: New Rights of Purification? The handling of political corruption scandals and the Japanese coalition governments since 1993; Steven R. Reed, Chuo University, Tokyo: Political reform and political money in Japan; David T. Johnson, University of Hawaii at Manoa: Why the Wicked Sleep: The prosecution of political corruption in post-war Japan; moreover I gratefully acknowledge points made by my fellow-commentator Arnold Heidenheimer, University of Washington, St. Louis, and from the audience. Ms. Dagmar Lee, Mr.Christian Lder and Mr. Frank Robaschik of Duisburg University have helped me with collecting and editing the material.

to escape from the rather narrow interpretations of events and developments which are based on a peculiar tradition of research, which may sometimes lead into analytical dead ends. Apart, Japanese Studies should try to incorporate a comparative perspective in order to help avoid a point of view which too easily presupposes idiosyncratic national characteristics - without noticing that in other countries developments may be quite similar and, therefore and somewhat ironically, what is truly peculiar about Japan.

The paper is organised as follows: After a brief definition of corruption, an overview on how corruption developed in Japan is given. This is the basis for the following more analytical and abstract parts of the paper. In chapter 3, I will discuss the effects of corruption, using some rather standard economic techniques. In the following part, the focus is on possible changes in corruption and its effects during recent years. Finally, I will discuss some policy aspects of reducing corruption.

2. Corruption in Japan An Overview

For this discussion, corruption can quite simply be understood as the use of public office for private gains. Sometimes, a distinction between political corruption and other types of corruption is made. For instance, the term economic corruption is used when the private gains take the form of monetary rewards, whereas in political corruption they consist of raising the chances of re-election or remaining in office. However, such a distinction is of limited value, because gains can indeed take a different form and it is not obvious whether this difference really has significant effects. Another way of distinguishing between the various types of corruption refers to the individual who is being corrupted. From this point of view one might call a case in which a politician is corrupted political corruption, in case of a bureaucrat it is bureaucratic corruption and in case of private business (e.g. bribing a real estate agent to make a beneficial deal) it is business of economic corruption. For our purposes, it is sufficient to state that we are concerned with political and bureaucratic corruption in the latter sense and that we do not strictly distinguish between what form the private gains take.

Based on this definition, to what extent do we find corruption in Japan? In recent years, the frequency of corruption scandals in Japan has certainly increased. To gain an overview, a list of arguably the most serious corruption scandals of the post-war period is presented below. This list is by no means exhaustive (Imidas 1995, 1999, Blechinger 1998):

The Shipbuilding Scandal of 1954 In 1954, it was uncovered that to lobby for a 1953 law which allowed shipbuilding companies to borrow below the market rate, huge bribes were paid to politicians and leading bureaucrats. The scandal contributed to the collapse of the Yoshida cabinet, but only one person convicted was actually sent to prison, out of 71 persons arrested. Sato Eisaku, who was also involved, would later become Prime Minister (1964-72) and receive the Nobel Prize in 1974.

The Lockheed Scandal of 1976 At the summit meeting between President Nixon and Japanese Prime Minister Tanaka in August 1972 it was agreed that Japan was to import large numbers of Lockheed aircraft. In July 1976, i.e. when the Miki government was in power, Tanaka was arrested for having allegedly received so-called success payments from Lockheed, amounting to the sum of 500 million Yen. Altogether some 460 persons, among them 17 members of parliament, were questioned in connection with this affair. Although there was proof that monies had been accepted, no charges were made. Instead, the names of the 17 members of parliament were published. Both in the first and second court hearing in 1983 and 1987 respectively, Tanaka was sentenced. He submitted an appeal, but with his death in 1993 court proceedings came to an end.

The Recruit Scandal of 1988-89 The Recruit Scandal centred around Recruit Cosmos, a part of the Recruit Group of Enterprises, accused of issuing shares to leading politicians, civil servants, and representatives of associations and the mass media, before these were traded at the stock exchange. The idea was to enable the persons thus favoured to sell the shares again with profit after they were officially listed. The financing of these deals was often effected

through interest-free loans made available by one of Recruits own finance companies. In exchange for these transactions and political donations, various enterprises belonging to the Recruit Group received substantial help from politicians and the bureaucracy, thus gaining considerable advantages. The scandal made clear that not only had indvidual politicians been bribed, but almost all important politicians had accepted payments from Recruit. The affair came to a head with the resignation of Prime Minister Takeshita in April 1989. It came to light that, among others, former Prime Minister Nakasone and a member of his cabinet, Mr. Fujinami, were also heavily involved.

The Kywa Affair of 1991 This affair involved payments to Abe Fumio by Kywa, a steel-girder construction firm. When the scandal broke, Abe was Secretary General of the Miyazawa faction of the LDP. Prior to that, he had been head of the Hokkaid and Okinawa Development Agencies. In exchange for payments, Abe arranged (via another politician) contacts to the Marubeni trading company, providing Kywa by also bringing in former Prime Minister Suzuki with the approval to build a golf course. Amid accusations of corruption, Abe resigned in December 1991. He was arrested in January 1992 and in May 1994 sentenced to two years imprisonment.

The Sagawa Kybin Scandal of 1991-1993 The Sagawa Kybin parcel service firm donated generous sums of money to politicians of the LDP responsible for transport matters as well as to other influential politicians in other parties. A rapidly expanding firm, Sagawa Kybin had high hopes of thus obtaining the licences for a nationwide parcel service. What was special about this affair was the fact that monies had not only been paid to politicians, but also to one of the syndicates of organised crime (Yakuza). The fact that Kanemaru Shin, the Deputy Secretary General of the LDP, had actively sought such contacts while engaged in the election campaign of Takeshita Noboru, greatly damaged confidence in the LDP.

The Tax Evasion Scandal of 1993 surrounding Kanemaru Shin While the other scandals involved the financing of political activities, this was arguably a case of personal enrichment and thus rather different, especially when it emerged that the

extent of the tax evasion was quite enormous. During a house search valuables worth 3,600 million Yen were seized.

The Genecon Corruption Scandal of 1993 In connection with payments to politicians by large enterprises of the building and construction industry (general contractors genecon), the mayor of the city of Sendai was arrested in 1993. The scandal soon widened, involving also the governors of the Prefectures of Ibaraki and Miyagi. In October 1997, Nakamura Kishir, the former Minister for Construction, was convicted of corruption.

The Skaiya Scandals of 1997 This involved four large broking firms (Nomura Shken, Yamaichi Shken, Nikk Shken and Daiwa Shken) and the Daiichi Kangy Bank, accused of paying monies to persons blackmailing the management in connection with the shareholders meeting.

The Scandals of 1996-1998 within the Elite Bureaucracy In the second half of the nineties a number of scandals within the upper echelons of the bureaucracy were uncovered. The bribery of a high official of the Ministery of Health and Welfare who had accepted payments and hospitality by a firm engaged in the building of old peoples homes hit the headlines together with the excessive offering of hospitality to some highly placed staff of the Bank of Japan and to inspectors of the Ministry of Finance, who, in exchange, passed on confidential information about the planned inspection of credit institutions by the supervisory authority.

Does this rather long list of only the major scandals suggest that there has been an increased propensity for corruption in Japan in recent years? This conclusion would certainly be premature, because one lacks reliable data on how much corruption really takes place, how many people, how many cases, which monetary value and which number of votes bought are involved. It cannot be assumed that there exists a linear relationship between the number of scandals which are reported in the media and followed up through the judicial process and the level of unobserved corruption. Such a relationship cannot easily be established, because the reason why cases of corruption are made public may have to do with finding scapegoats,

upheaval of traditional relationships etc. So quite paradoxically, an increased number of scandals can both signify that corruption has increased as it may signify that considerable efforts are undertaken to stop corruption happening.

Another way to handle this difficult issue is to engage in careful historical analysis. Based on such an analysis, Mitchell 1996 can show that Japan has a history of corruption dating back at least to the Heian period; repeated traces of corruption can be found more than 1000 years ago.

Based on such observations one might easily be inclined to propose that Japan possesses an intrinsic orientation towards corruption which may be rooted in its cultural tissue. However, if one engages in comparative analysis across countries and time, it can easily be shown that instances of corruption have occurred in all countries and at all times, at least dating back to ancient Rome, Greece and China. Considerable efforts have been made to rank countries numerically in terms of their propensity for corruption. As the underlying cases of corruption or their volume cannot be used for such an analysis as has been pointed out above, one usually relies on subjective evaluations by informed observers, which through careful selection and interpretation possess some degree of objectivity. In Table 1, two of such surveys are reported. The Business International (BI) Survey is based on correspondents reports in about 70 countries. It has recently been incorporated into the Economist Intelligence Unit. BI staff try to ensure accuracy and consistency of the evaluations. It is understood that the BI index may turn out to be somewhat biased in terms of measuring business corruption, and there is the danger to be reckoned with that correspondents are influenced by overall economic success and achieved per capita income of the respective countries. As for Transparency International, it is a meta study of at least some six or seven underlying surveys, increasing its reliability. The variance of the scores is usually within about 1 index point for industrialised economies. However, in the case of Japan, with a mean score of 7.1, even the variance reaches 2.61. This shows that different surveys come to a rather wide range of conclusions. Looking at Table 1 more closely, one notices that among leading industrialised economies, Italy and Japan fare rather badly. Moreover, the Transparency International values for 1996 are in both cases considerably lower than the Business International figures based on an average for 1982 1993; this holds for other countries as well, though. Although the data looks hardly

trustworthy enough to engage in sophisticated quantitative analysis based on it, it does at least show that, from a comparative point of view, Japan is not necessarily an outsider. Japan just seems to experience its problems of corruption like other industrialised economies as well. It may have some more of it than others (particularly Australia, which very often together with New Zealand and Scandinavian economies have the best scores), but it is not special in its concerns.

Table 1 Indices of corruption for selected countries

Business International Survey*

Transparency International Survey*

Australia Brazil France Germany Italy Japan Korea (RoK) Russia UK US

10 5.75 10 9.5 7.5 8.75 5.75 n.a. 9.25 10

8.6 3.0 7.0 8.3 3.4 7.1 5.0 2.6 8.4 7.7

Sources: Mauro 1995 for BI, Bardhan 1997 for TI * Refers to an average of observations between 1980 and 1993. Ranges from 10 (good) to 0 (bad).

** Average of several studies (meta-study) for 1996. Ranges from 10 (good) to 0 (bad).

3. The Effects of Corruption

It is usually rather plainly taken for granted that the effects of corruption (on the economy) are negative. Basically, of course, this argument is right, and an economist would argue that two facts in particular are responsible for such negative effects:

Through corruption, the optimal allocation of resources is impeded. Such resources are redirected into less efficient utilisation than compared to a situation in which the price mechanism based on the rules of a well organised market economy can do its job.

Corruption has a detrimental effect on investment, entrepreneurship, and innovation. In case entrepreneurs have to fear that the future is to some extent shaped by corruption, possibly from their adversaries, this raises their risk associated with economic activity and will lower risk taking entrepreneurial behaviour. Of course, this has severe negative effects on economic growth.

Economists have raised a few points against these seemingly straightforward arguments, though, which complicate the picture:

The first argument is taken from the economics of crime (Kunz 1993). Fighting all corruption will finally reduce the number of bads originating from corruption to zero. On the other hand, the cost associated with fighting corruption will also increase. Trying to repress even the last instances of corruption may possibly be very expensive indeed. In consequence, there will be a point between zero percent and one hundred percent of corruption, at which the additional benefits of avoiding one more instance of corruption just outweigh the (upward sloping) cost of anti-corruption activities. From an economic point of view, it is therefore not sensible to try to get rid of all corruption. Of course, one may doubt whether it would be possible at all to get rid of all corruption, even if one wanted to; translated into the terminology of economics, the (marginal) costs of fighting the final percentage points of corruption may indeed be prohibitive.

We have so far assumed that getting rid of an incident of corruption does indeed raise social benefits. Some economists have pointed out, however, that corruption may actually have some positive effects, namely when regulation in the economy is so rigid that economic dynamism is severely throttled. Corruption and bribery may then serve as grease for the economy. Samuel P. Huntington takes a similar position when he states: In terms of economic growth, the only thing worse than a society with rigid, overcentralised, dishonest bureaucracy is one with a rigid, overcentralised, honest bureaucracy (1968, p. 386, quoted from Bardhan 1997, p. 1322). While deregulation would still be the first best solution in such circumstances, corruption may turn out to be second best. One instance is speed money. In case of overregulation, paying speed money to bureaucrats may be the best alternative to get important economic activity started.

Despite these points which seem to raise considerable doubt about the seemingly simple relationship between corruption and its detrimental effects on economic development, economists in recent years have repeatedly and consistently argued that overall the negative aspects of corruption will nevertheless be greater. As evidence, rather sophisticated quantitative studies are used, although based on the limitations of the data, they may not be considered fully convincing to any reader (e.g. Mauro 1995).

It is therefore sensible to have a look at some more analytical arguments. A first point is that those economic activities favoured by corruption in a heavily regulated environment are not necessarily those with the highest positive contribution to allocate efficiency or investment and growth. Rent seeking activities may promise higher rewards and may therefore be more prominent than productive economic activities in which a net value-added is gained. Another point is that in a regulated environment with corruption, politicians and bureaucrats have a strong incentive to introduce even more regulation in order to gain more chances of collecting bribes. This implies that the regulative level is no exogenous variable, but may possess selfreinforcing properties if combined with corruption which have even more negative effects.

Summing up, while it is still true that corruption is an impediment to allocation and growth, careful economic analysis based on rather standard methods shows that the relationships involved are actually somewhat more complicated than often presumed.

What implications does this hold for Japan? There is a simple fact about the Japanese development, which seems rather difficult to explain, but on which the ideas just discussed contain an important message: The fact is that despite a rather significant level of corruption, probably for many centuries, this has not made economic development impossible. Indeed, despite structurally entrenched corruption, Japan as a late-starter has achieved enormous economic development. One may certainly oppose corruption in Japan, but would still be difficult to present substantial evidence that, for instance, the ship building scandal of 1964 or the Lockheed scandal of 1976 had a serious impact on the growth of the Japanese economy.

One argument put forward in this context is related to the indecisive effects of corruption in a regulated environment. It is pointed out that if corruption occurs as a one stop solution, the detrimental effects on allocation and growth may be somewhat limited. If companies know they can achieve a certain political decision or licensing agreement by paying a specific sum to a politician or bureaucrat, this helps them to make business decisions in a reliable framework and get on with their job. Of course, they may still have to face considerable transaction costs in contacting the politicians or bureaucrats and of trying to bargain with them (always risking to be exposed), however, once the deal is struck, the bribe is just an additional fixed cost on the project. In other countries, which are characterised by decentralised corruption, the distortionary facts may be much higher (Schleifer and Vishny 1993). Companies may try to bribe various officials and organisations to achieve a certain outcome, but they will still not be sure whether they have really approached the right people or whether another institution or person may also want to be bribed and has the means to stop the whole project until he is satisfied. Given similar levels of overall corruption, the result of corruption may thus be rather different, depending on its structure.

Arguably, Japan is a typical case of one stop corruption. The notorious so-called iron triangle may be interpreted as a rather dense network in which an o.k. by a leading politician and his followers justifies the expectation that the consent of this group of people is a


trustworthy commitment on their part and, that they possess the means to do away with any further obstacles.

This argument may not only be relevant for Japan, but also for other countries of East Asia. Pranab Bardhan notes: The strong states of East Asia with their centralized rent-collection machinery and their dense encompassing network with business interests stand in sharp contrast [e.g. to Africa, W. P.], even though by some measures corruption has been quite substantial. ... the ability to precommit credibly may have been an important feature of the strength of such states (1997, p. 1341). I have quoted this section because it is from the final, and thus important paragraph of a review article on corruption and development in a leading economics journal, namely the Journal of Economic Literature of the American Economic Association. This shows that East Asia constitutes an important case for studies of corruption, and Area Studies has a lot to contribute to the disciplinary and inter-disciplinary discourse on corruption.

4. Changes A New Role for Corruption in the Japan of the 1990s?

The discussion should now turn to the question, whether significant changes of corruption can be found in Japan. In many evaluations, including those which formed the background on which this comment was produced (see Footnote*), it is proposed that although change may be somewhat slow, the 1990s have witnessed a prolonged crackdown on corruption and there is a glimpse of hope that corruption will retreat. Of course, few would be optimistic enough to argue that all corruption could ever be abolished whether in Japan or elsewhere.

The usual arguments given in this context are related to:

an increased transparency of the political and bureaucratic process as a reaction to the crisis of the political and bureaucratic establishment

electoral reform with a structurally increased role for the parties and their programmes and a decreased role for the support groups of individual candidates


an enhanced accountability of politicians in the legal process (for instance, responsibility of politicians for actions of their political secretaries or other aides)

a growing frustration of big business with a corrupt and to some extent inefficient political and bureaucratic process.

I will not enter into a deeper analysis of these ad-hoc factors. Rather, I would like to add some possibly rather abstract, but nevertheless noteworthy considerations from an economic perspective.

The first point to be made is that if corruption did indeed not have overwhelming negative effects on the Japanese economy under the system of the iron triangle and the one stop character of corruption in Japan, then it is indeed questionable whether a presumed decline of corruption has indeed net beneficial effects on the Japanese economy. In particular, these beneficial effects could be overcompensated by the negative effects of one stop corrupt situations being reduced: the relations among politicians, bureaucrats and business have become more insecure and unpredictable in recent years and this may have worsened possible effects of corruption for the time being.

A second point relates to the question whether some circumstances in the environment or the border conditions of corruption (transparency, electoral law etc.) do indeed change the propensity for corruption behaviour significantly. There is some disturbing economic analysis which suggests that it may be very difficult to get away from a socio-economic system in involving high or low degrees of corruption: Although this suggestion may be surprising, it can be rather easily derived from a simple model. Suppose that we draw a graph in which the horizontal axis shows the proportion of corrupt politicians (or bureaucrats) in the total population (Figure 1). At the very left side there are no corrupt individuals whereas on the right hand side everybody is corrupt. Two curves are drawn in this setting. The first curve including points B and C denotes the advantage of being corrupt. At the left side, the advantage is quite low because the possibility of being caught is high. With an increasing overall number of corrupt colleagues, this danger declines and marginal benefits of being corrupt increase until there is a point when the advantages decline again because so many other persons are corrupt that bribes are competed away. There is another curve which includes


points A, B and D. It shows the advantage of being honest. When everyone else is honest, the advantages are quite high (Point A), but the more dishonest colleagues there are, the more the advantages of being honest oneself decline. At the right end of the graph, one may even possess considerable disadvantages (Point D) at being the only honest official left. There is an equilibrium in Point B. At this point it is just as advantageous to be corrupt as it is to be honest; so there is no incentive to change. However, if due to changes in the circumstances there a few colleagues turning corrupt or non-corrupt (i.e. a movement somewhat to the right or to the left from Point B) there is a net advantage to be corrupt (on the right hand side) or to be honest (on the other side) for oneself as well. In economic terminology, B is therefore an unstable equilibrium and even with minor disturbances, the whole system will move either to the very left or to the very right side. Given any kind of disturbance near the edges, the system will move back (in a stable manner) to either O or N. In other words: If we are in a heavily corrupted system (close to N), even rather determined anti-corruption policies to move left will only lead to a later backlash unless we can move so far to reach point B, but that may be extremely difficult and unlikely.

Figure 1: The incidence of corruption as a multiple equilibria problem marginal benefits A B advantage of being corrupt

C advantage of being honest

O no one is corrupt

unstable equilibrium

N D all are corrupt

Source: Bardham 1997, p. 1331, with minor changes


Despite this model being extremely simple, it already has properties of a multi equilibria system in which it is very difficult to move away either from an honest or from a corrupt politicoeconomic system. The literature on the role of institutions has recently been full of assertions to such multi equilibria situations (e.g. Arrow 1998 with application to East Asia) and corruption may be one case in point.

As for Japan, this may imply that despite the recent changes in terms of electoral reform etc. the whole system may prove extremely slow in moving away from a corruption environment. Given the decline of one stop corruption as well, one may indeed be sceptical whether the overall impact of the recent changes can indeed be considered to be positive.

It is important to note that based on the model presented above, nations will turn towards the stable ends of the spectrum (high/low corruption) even without differences in political or economic culture. Does the model show that culture definitely does not play a role in bringing about corruption, so that all analyses based on studying corruption from a cultural point of view are flawed? Certainly not. The first aspect to be taken into consideration is that the above mentioned model only shows that a pervasive corruption environment may come about even without important cultural peculiarities. It does not contain any suggestion about the possibility that cultural backgrounds can as well be important.

Secondly, and possibly somewhat more interestingly, one may question whether the model can sensibly be applied to Japan at all. The graph presented (Fig. 1) contains the underlying assumption that corruption spreads (or retreats) evenly through the economy, i.e. corruption behaviour in policy arena X has a direct impact on such patterns of behaviour to occur in arena Y. For instance, in the case of customs officials being corrupt, civil servants engaged in regulating the construction industry will take that as an important signal for themselves about which marginal benefits and costs (or risks) they may encounter when they act honestly or corruptly. When a large number of customs officials are corrupt, regulators of the construction industry will suppose that this corrupt behaviour is typical of their environment as well and act accordingly. However, in economies which are grouporiented in the sense of many relationships being based on long-term (implicit or explicit) contracts between specific, noninterchangeable individuals, this condition must not necessarily hold. Many observers have


argued, that indeed in many policy arenas Japans socio-economic system is indeed characterised by such long-term network or generalized exchange relationships (e.g. Pascha 1996).

If we assume such an environment, an honest or corrupt situation in arena X will not necessarily be of any consequence for arena Y. This holds because the individual actors of arena Y only take notice of other actors of arena Y, because these are the only ones they are related to, disregarding other arenas X or Z. For Japan, this implies that there may be arenas with very high levels of corruption (e.g. in electoral politics), while in other areas there may be an extremely low propensity for corruption (try to bribe a Japanese customs official to find out!).

Put differently, the model presented in Figure 1 has been shown to be culture-specific. It supposes an environment characterised by rather short-term anonymous (balanced exchange) relationships. Within it, the assumption of how corrupt the average citizen (or official) is expected to be, is the only information available and will therefore influence everybodys behaviour, easily leading to lock-in situations.

What does this imply for Japan, when we take structural changes in the Japanese socioeconomy into consideration? We now have one critical factor to look for in order to answer this question, namely to what extent there is a change between long-term and short-term (generalized and balanced exchange) relationships. This issue can only be sketchily discussed here, but there is some evidence that in various spheres of life long-term relationships decline while more anonymous short-term relationship do increase. This is related to the changing environment of social relationships, for instance, the benefits and costs associated with leaving existing relationships (chances of exit) (see Pascha 1996). If exit is easier in modern Japanese society, we can derive the hypothesis that the hitherto rather isolated policy arenas with their peculiar relative incidence of honesty or corruption will develop more links in the sense that behavioural expectations can more easily cross arena boundaries. This holds because dense long-term relationships, cementing a specific arena, decline, whereas unstable short-term relationships increase and develop into many directions. As a simple illustration, we may observe more behaviour of the type of a customs official thinking: If high-ranking politicians


and bureaucrats are as corrupt as I can read in the papers, why should I be non-corruptible given my small income, shabby government-supplied housing etc.? Given the multi equilibrium properties of the spread of corruption as noted in the model underlying Fig. 1, this may have surprising and disturbing dynamic properties.

Summing up: While many observers believe that electoral reform, more transparency in the political process etc. lead to less corruption in Japan, we have found evidence that there may be two factors which seriously work against such an outcome:

The move away from one stop corruption may lead to a more insecure, unpredictable politico-economic environment, in which corruption may actually have more negative impact on development and growth than before.

As behavioural boundaries between various policy arenas slowly lose importance, even arenas in which corruption hitherto has not played any significant role may experience it, and this may counteract some positive tendencies in other areas.

5. Policy Issues

We now turn to the final point on adequate policies to fight corruption. From an economists point of view, the basic problem is that politicians or bureaucrats have an almost natural incentive to abuse power conveyed by the society for public purposes. This problem can also be stated in the terminology of the so-called principal-agent problem of economics (for a pioneering application to corruption, see Rose-Ackerman 1978). Citizens or voters can be seen as the principals who want certain public policy instruments with peculiar objectives to be applied. As the principal is not able to handle it herself, agents (politicians, bureaucrats) are used. The problem arises how to make sure that the agents really act in the interest of the principal. This is particularly tricky, because there will usually be an asymmetry of information. Agents will know better which conditions they face and which options they have, while the principal will have difficulties to judge whether the agent has acted honestly. Fighting corruption therefore typically leads to considering chains of principal-agent relationships, containing at least one intermediate supervisory level, and possibly more. For instance, one


may have a first stage of supervision actually monitoring the street-level bureaucrat, whereas in a second stage the legal status of the process at hand is reviewed. One may discuss such a framework with respect to its effect on external corruption, i.e. on possible collusion between the street-level bureaucrat and the client, or on internal corruption, i.e. on collusion between street-level bureaucrats and supervisors (see for instance Bac 1998).

Unfortunately, there is no simple rule to an optimal institutional design for a workable principal-supervisor-agent framework. I consider it at least doubtful, though, whether the simple information that politicians in recent years have played an increasing role in trying to interfere with the elite bureaucracy of the ministries in itself already constitutes a superior solution to the earlier set-up. Politicians also act in their self-interest and given the information asymmetry which severely handicaps the electorate as the principal in charge, little optimism seems to be warranted.

Quite a few policy proposals in recent years seem to be based on the idea that an independent agency is needed to supervise street-level politicians or bureaucrats, with incentives and rules of such agencies finely tuned to the goal to uncover and prosecute corruption. This independent agency issue incidentally is quite similar in various policy issues of Japans political economy, not only with respect to corruption:

In monetary policy, there is the discussion of whether the newly independent Bank of Japan can make a fruitful contribution to macro-economic development in Japan, better than under the earlier framework of its somewhat intransparent subordination under the Ministry of Finance.

Also with respect to the Ministry of Finance, the set-up of an independent financial supervisory agency is much discussed and closely watched. Will it effectively and efficiently be able to take care of a sound financial system?

Moreover it is being discussed whether a more independent Fair Trade Commission with the role and influence of MITI decreased will be able to achieve superior results in terms of safeguarding workable competition.


Discussing the possibility and effects of independent agencies within a Japanese socioeconomic framework, is an extremely interesting cross-cutting issue. This can only be introduced here as a question for further research which may shed further light on various policy issues of todays Japan.

References Arrow, Kenneth J. (1998) The Place of Institutions in the Economy: A Theoretical Perspective, in: Yujiro Hayami and Masahiko Aoki (eds.) The Institutional Foundations of East Asian Economic Development, Basingstoke et al.: Macmillan, pp. 39-48 Bac, Mehmet (1998) The Scope, Timing and Type of Corruption, International Review of Law and Economics, 18, pp. 101-120 Bardhan, Pranab (1997) Corruption and Development: A Review of Issues, Journal of Economic Literature, Vol. XXXV, September, pp. 1320-1346 Blechinger, Verena (1998) Politische Korruption in Japan. Ursachen, Hintergrnde und Reformversuche, Hamburg: Institut fr Asienkunde Hirose, Michisada (1993) Seiji to kane (Politics and money), Tokyo: Iwanami Huntington , Samuel P. (1968) Political Order in Changing Societies, New Haven: Yale University Press Imidas (various years), Tokyo: Sheisha Kunz, Harald (1993) Kriminalitt, in: Bernd-Thomas Ramb, Manfred Tietzel (eds.) konomische Verhaltenstheorie, Mnchen: Vahlen, pp. 181-206 Mauro, Paolo (1995) Corruption and Growth, The Quarterly Journal of Economics, August, pp. 681-712 Mitchell, Richard H. (1996) Political Bribery in Japan, Honolulu: University of Hawaii Press Pascha, Werner (1996) On the Analysis of Change and Continuity in Japans Socio-Economy, in: Sarah Metzger-Court and Werner Pascha (eds.) Japans Socio-Economic Evolution. Continuity and Change, Folkestone: Japan Library, pp. 27-49 Rose-Ackerman, Susan (1978) Corruption. A Study in Political Economy, New York et al.: Academic Press


Shleifer, Andrei and Robert W. Vishny (1993) Corruption, The Quarterly Journal of Economics, August, pp. 599-617